Cluster interpretations and/or observations

On the previous page, you made observations about patterns in your data. For each observation, you entered multiple interpretations about what that observation might mean (to reasonable people who happen to disagree).

On this page you can draw your observations and/or interpretations together into meaningful clusters. Clustering creates a two-tiered structure to your catalysis report. People will be able to look at your report at a high level, then "drill down" to the details that lie behind each cluster.

What to do here

First choose a catalysis report to work on. Then, for either interpretations of observations (or both), click the Copy or update button above the surface you want to work on. Move the items around, placing like with like.

Shift-click to select multiple items. You can also draw a "rubber band" box to select multiple items. Shift-drag to add more items to the selection. Click anywhere outside an item to deselect everything.

When you are satisfied that all of your interpretations/observations are in groups that make sense, name the groups. For each group, choose Create new cluster name from the list of things you can do (under the clustering space), then click Do it. Type in a name and description, then click OK. Once you have named and described all of your clusters, you are ready to print your catalysis report. Your cluster names will become the headings of your report.

The headings in a catalysis report with clustered interpretations are called "perspectives." The headings in a report with clustered observations are called "themes."

Connections to other pages

In order to use this page, you must first create a catalysis report (on the Start catalysis report page) and add some observations and/or interpretations to it (on the Explore patterns page). The way you cluster your observations and/or interpretations on this page will affect the way your catalysis report appears (on the Print catalysis report page).

Frequently-asked questions

What do the circle colors mean?

Gray: an interpretation, or an observation with no strength value set
Orange: an observation with a strength value set (the brighter the orange the stronger the observation)
Blue: a cluster name
White: an item or cluster you have hidden (it will not appear in your report)

Why do I see an interpretation on this page that I cannot see on the "Explore patterns" page?

Your choice of graph types and questions at the top of the "Configure catalysis report" page has no effect on your printed report. It only controls what you can see on the "Explore patterns" page. Every interpretation you write is automatically included in the catalysis report - that is, unless it is associated with an observation with a strength value you chose not to print.

To exclude a particular interpretation from your report, hide it in the relevant clustering space. In the space, select the interpretation circle, pick "Show/hide selected item/cluster in report" from the list below the space, and click "Do it". Hidden interpretations (and observations and clusters) are drawn as white circles.

Should I cluster interpretations or observations?

Perspectives and themes organize your catalysis report in different ways.

Perspectives - overall points of view
- Interpretations - what some people might think
  - Observations - what those thoughts could be based on
Themes - overall trends
- Observations - patterns anyone can see
  - Multiple interpretations - what each pattern means to people with different perspectives

Which cluster type you should use depends on how you are going to use the report and who is going to work with it. Clustering by interpretations gives strong attention to multiple points of view on the stories you collected. If you are using the report in a facilitated workshop, and you want to challenge people to think differently, clustering by interpretation can be a great way to break down assumptions and get people thinking in new ways. This can also be a useful method of organization when the topic is sensitive and people are likely to be wary of "scientific" approaches to understanding it.

On the other hand, a clustered-interpretations report is longer and somewhat more confusing, especially since observations are repeated after each interpretation they are connected to.

If your ambitions are more modest, or if you think people will be less than enthusiastic about exploring multiple perspectives, or if you just have a lot of information for people to get through, you might want to group your report by clustering observations. People will still see multiple interpretations for each observation, but they'll see them after the observation, which might make them seem less "real" or important. (And that might bother some people.)

You don't have to choose only one form of clustering. You can cluster your observations and your interpretations. Each type of clustering provides a different type of summary and a different type of insight. By clustering both ways, you can print both types of report. Then you can use each report in a different situation, or for different groups who want to work with the patterns differently.